Doubling our impact with Atlassian Cloud

[Stephen Neil]

Hi everyone 

This is my first time posting here, so a quick hello — I’m Stephen and I lead on IT Infrastructure & Operations at Mary’s Meals International. We run a global school feeding programme, currently reaching over 2.6 million children across 16 countries every school day.

We recently shared a case study with the Atlassian Foundation on how we’ve built a global System of Work using Jira, Jira Service Management, Confluence, Bitbucket, and Statuspage, helping 21 international teams collaborate on everything from IT support to marketing, HR, travel, and legal ops.

Since migrating to Atlassian Cloud last year, we’ve particularly appreciated:

• Reduced time and effort around upgrades and maintenance

• The improved performance and scalability available to us

• The value of Atlassian Intelligence for improving response times & multi-language comms, drafting epics/stories/sub-tasks and suggesting Knowledge Base content.

Happy to connect or answer questions about how we’ve approached Cloud migration, working with affiliates / federated orgs, or anything else that's helpful.

Here’s the case study

Thanks to the Atlassian Foundation for the support!

[Emma Schnee]

It's been great collaborating with you, @Stephen Neil ! Thanks for sharing here :)

[Fun Man Andy]

This is so cool @Stephen Neil !! 😎 👏 

That success over 22 yr of 300 meals to millions of mouths fed is a dream.

In have some cool questions for you:

1. How many months in total did it take to migrate from SC to Cloud!?
2. What was the number advantage for systems maintenance moving to cloud.
3. Any chance in the future that your team (or even the whole org!!) would reach 100% SLA Compliance, and do you think Rovo will be the key to unlocking that potential?
4. Have you looked into the Teamwork Collection with Atlassian, and if yes, What are your thoughts on that?

[Stephen Neil]

Hi Andy,

Thanks! It’s been a fantastic journey to be part of.

On your questions:

Migration timeline – From kicking off planning to being fully live in Cloud took us about 4-5 months. The migration itself was done over a weekend — most of the time was preparation, testing, and making sure all teams were ready. Using an Atlassian partner was key for us.

Maintenance advantage – The biggest win is the time we no longer spend on upgrades, patching, and testing (particularly integrations, which we've reduced on cloud). I'd estimate this to be at least a man day a month on average. We had a partner managed DC system before but were still involved in testing. The time saved (both from us and our partner) can now go into improving workflows and supporting teams.

100% SLA compliance – It’s always something to aim for! The reality is some requests depend on other teams or external factors, and some of our team members are field based, travelling in very rural and disconnected communities. 100% would be tricky — but I am hopeful that AI features like Rovo will help us handle the predictable, repeatable requests faster so we can focus on the more complex ones. We've yet to unleash the portal agents but I'm keen to and we're doing some testing later this quarter.

Teamwork Collection – Not something I've looked into. We’ve already got our toolset in place using some of the components (not Loom, and Rovo only early stages). I can see the appeal for newer teams.

Cheers,
Stephen

[Josh Costella]

Hi @Stephen Neil 

This is such a great use case for an even greater purpose! So glad to hear that the Atlassian ecosystem helped you and your teams scale successfully. 

How did you address things like onboarding users into the tools and ensuring long-term adoption? Those can be challenging to accomplish when scaling. 

[Stephen Neil]

Hi Josh,

Thanks for the kind words — really appreciate it!

You’re absolutely right, onboarding and long-term adoption are some of the trickiest bits when scaling, especially across a mix of countries, roles, and digital maturity.

Thinking of a few things helped us:

Early involvement – We brought people into the process early (both the migration and the development of new projects using the tools) so they could shape how those tools and workflows would work in their context. That gives us more champions and less resistance.

Champions – Still developing this area but we try to identify someone in each team to train/coach/support others — that makes a difference, especially when time zones or language could’ve been barriers.

Mission-first framing – Instead of “here’s a new tool,” we talk about how the tools help us reach more children — which is something everyone here is deeply committed to.

Iterative rollout – We didn’t try to do everything at once. We started small, showed wins, and kept adapting based on feedback. Our first JSM project was the IT Service Desk and our first Jira Software project was managing the rollout of our integrated digital platform (CMS + CRM + Marketing Platform). These projects touched a lot of our org and showed people how it could work. Showing early improvements in things like request turnaround or onboarding experience built trust. Other teams then began asking how they could utilise the same tools themselves in their areas. There's many more functions Id love to onboard. As time allows I build them PoC projects to show and tell potential value.

Sustained adoption is still an active effort, but that foundation helped us get off to a strong start.

[carolyn french]

@Stephen Neil - I love what you've done! And thanks for sharing how you're using Rovo too.

Did you run into any roadblocks with the Cloud migration?

[Stephen Neil]

Hi Carolyn,

thanks for the encouragement. Yes — we did hit a few bumps in the road during the Cloud migration. Most were manageable, but they taught us a lot. 

• Azure AD group syncing caused early delays in access provisioning — had to get that stable before UAT could start. (moving from Crowd to Atlassian Access)
• Not all assets migrated cleanly (we focused on laptops first and sorted the rest post-migration). It's a shame there's not a tool that migrates asset history and comments - we manually reviewed and copied over what we thought we'd like to keep.
• Broken Links - on DC we'd used custom domains. With no custom aliases for Confluence on Standard plan this meant broken links and a rethink on how we deep-link docs. At the time this wasn't something the migration tools could help with, though support advised on a programmatic fix.
• A couple of apps didn’t fully migrate — like the Microsoft 365 for Jira app, which got stuck at ~97%. We didn't actually bottom out why. This was possibly behind some of the permission issues we observed when trying to access Teams chat histories linked to tickets. We focussed on fixing the permissions for the "open" tickets that were being migrated.
• And staging sent out real notifications, which caused some momentary confusion (“why is this ticket closing?!”).

Most of these were small and fixable, but all good reminders of why UAT matters. We kept a detailed issue tracker and fixed what mattered most.

A couple of things we flagged as “won’t fix” because they weren’t blockers — just legacy differences we could live with or adapt to post-migration (e.g. some custom fields and board views didn’t behave quite as expected — things like epic names vs summaries showing differently).

The structured approach that our Atlassian partner took for the migration and UAT was invaluable.